The Falcon 9 (1.0 and 1.1 combined) has had one partial failure and 12 successful launches, the Antares one complete failure out of five launches, the Delta II one failure (and one partial failure) out of 152, the Delta IV Medium 20 successful launches with no failures, the IV heavy 7 successes and 1 partial failure on a test flight, the Atlas V 51 successes and 1 partial failure. Yes, the Delta III was horrible, but it was only launched three times back in the 90s and abandoned.
The DoD launch you're talking about happened in 2007. No other US company could get a satellite in GSO at the time. SpaceX had only launched two Falcon 1s for DARPA at that point, both too small, and both failures. Orbital at least had their Pegasus... with ~1/10th the required payload and a poor success rate.
I'm not a ULA apologist, they were simply the only game in town for US satellite launches, and charged accordingly. SpaceX's recent successes have put them on track to become serious competition, and that's great. But you'd be crazy to trust a new space company with high-value payloads until they have a few successful launches under their belt.
There are only three US companies with LEO capabilities: ULA (Boeing/Lockheed), SpaceX, and Orbital.
Of those three, Orbital's Antares is currently grounded after its spontaneous disassembly a few months ago, and our darling SpaceX's Falcon 9 1.1 has only been in use since 2013. ULA's Delta and Atlas have longer and better track records and much higher payload capacity than the Antares or Falcon 9.
On top of that, SpaceX and Orbital have never handled classified payloads before, so that's training and time and effort on the part of the USAF.
While I wouldn't be surprised if there's some palm-greasing going on behind the scenes between USAF and ULA, I also can't blame them for not trusting startups with billion dollar spy satellites.
You're describing 'TurboCache' (a marketing name if ever there was one).
It wasn't a secret, it was only on very low end cards, and ATI was already doing the same with 'HyperMemory'. Intel, for their part, was exclusively using system RAM at the time (and largely still is).
So what graphics *have* you been buying for the last decade?
The problem with an EMP is you can't focus it.
I know what I'll be doing.
In fairness, it depends on what the passwords were *for*. If it's a bank site... that's bad. If it's some random site that hides content behind a pointless registration wall, '12345' is perfectly fine.
It comes down to 'if this were a door, would I lock it?'
Technically this is the FBI, so you should be pissed off at Comey, not Holder. Comey is officially Holder's subordinate at the DoJ, though I'm not sure how much the FBI chief really answers for.
And you won't have to wait so long for Holder's departure; he announced his resignation months ago and Obama already tapped his replacement.
Don't forget the Islamic State, which also counts oil as its primary source of income. Saudi Arabia has no shortage of reasons to drive down oil prices.
At this rate we'll need a metadata database metadatabase.
Less complexity, less weight (and gets lighter as you use it). No pumps, no power source for pumps, no return lines, just a pressurized tank and a few valves.
Of course, you have to know how much you'll need before the flight, and the longer you'll need it the lesser the savings over a traditional system.
Looks like most of my Kerbal Space Program landings.
edit: the above was supposed to be a reply to another post by gstoddart in this thread. It doesn't make as much sense in this context, sorry
"And remember: Evil will always prevail, because Good is dumb." -- Spaceballs